254 
The specimen no. 1278, collected by Henry in Ichang and 
determined by Oliver as P. fragiformis and also referred to in 
Index Florae Sinensis by Hemsley, is neither P. fragiformis nor 
P. Freyniana, but in fact P. fragaricides. Henry’s specimen 
collected in Yunnan (no. 10,9054), which is named P. fragiformis 
but not referred to in the Index, also represents P. fragarioides. 
h 
be quite reasonable. The plant has long been known as P. gelida 
amongst Japanese botanists. Th. Wolf cites Franchet and 
Savatier’s P. fragiformis var. gelida as one of the synonyms of 
P. Matsumurae. I rather fancy that the French authors meant 
P. Freyniana by this uame. efore the publication of his 
monograph Th. Wolf himself called the plant by the name of 
P : 
Th. Wolf published in his monograph a new variety of 
P. Freyniana called var. grandiflora (lc. p. 640). The same 
plant was described by Makino in the Tékyé Botanical Magazine 
(vol. xxiv, p. 142), with good reason as a new species :—P. Yoku- 
saiana. According to the Japanese author this plant has subpinnate 
radical leaves besides the ternate one. He separates this plant 
from P. Freyniana, pointing out as characteristic the absence of 
the thick rhizome and the peculiar form and serration of the leaflet. 
This last character often serves for classifying plants empirically, 
and usually is very difficult to define. 
The pinnate character of the radical leaf seems to me to be liable 
to variation in certain species. There is a specimen collected by 
Henry in Hupeh, which is not to be distinguished from P. fragarz- 
oides, although it possesses, like P. Freyniana, the ternate radical 
eaf; but by this feature alone we cannot recognise it as P. 
Freyniana, 
P. Dickinsit, Franch, et Sav., was reduced by Makino to 
P. ancistrifolia, Bunge, of North China. Koidzumi calls it var. 
Dickinsii of P. ancistrifolia. In his monograph Th. Wolf tries to 
distinguish these plants as two distinct species, especially by the 
colour and marking of the achene, the hairs on the receptacle and 
by the number and texture of the leaflets. As I have been unable 
leaflets is not constant. In P. Dickinsii the radical leaf varies 
from tri-jugous pinnate to simple with one rounded blade (forma 
